Revision 2 Texture, Geochemistry And

This is the peer-reviewed, final accepted version for American Mineralogist, published by the Mineralogical Society of America. The published version is subject to change. Cite as Authors (Year) Title. American Mineralogist, in press. DOI: https://doi.org/10.2138/am-2021-7889. http://www.minsocam.org/ P a g e | 1

1 Revision 2

3 Texture, geochemistry and geochronology

4 of titanite and pyrite: Fingerprint of

5 magmatic-hydrothermal fertile fluids in

6 the Jiaodong Au province

8 Xing-Hui Li1,2, Hong-Rui Fan1,2,3,*, Ri-Xiang Zhu2,3,4, Matthew Steele-

9 MacInnis5, Kui-Feng Yang1,2,3, Cai-Jie Liu6

11 1Key Laboratory of Mineral Resources, Institute of Geology and Geophysics,

12 Chinese Academy of Sciences, Beijing 100029, China

13 2Innovation Academy for Earth Science, Chinese Academy of Sciences, Beijing

14 100029, China

15 3College of Earth and Planetary Science, University of Chinese Academy of

16 Sciences, Beijing 100049, China

17 4State Key Laboratory of Lithospheric Evolution, Institute of Geology and

18 Geophysics, Chinese Academy of Sciences, Beijing 100029, China

19 5Department of Earth and Atmospheric Sciences, University of Alberta,

20 Edmonton, AB T6G 2E3, Canada

21 6No. 6 Institute of Geology and Mineral Resources Exploration of Shandong

22 Province, Zhaoyuan 265400, China

24 *Corresponding author. Tel.: +86 10 82998218; fax: +86 10 62010846. E-mail

25 address: [email protected] (H.-R., Fan).

Always consult and cite the final, published document. See http:/www.minsocam.org or GeoscienceWorld This is the peer-reviewed, final accepted version for American Mineralogist, published by the Mineralogical Society of America. The published version is subject to change. Cite as Authors (Year) Title. American Mineralogist, in press. DOI: https://doi.org/10.2138/am-2021-7889. http://www.minsocam.org/ P a g e | 2

26 ABSTRACT

27 The Au mineralization in the giant Jiaodong Au province is enigmatic and

28 difficult to fit current classic mineralization models, primarily because of

29 uncertainties as to the sources of ore-forming fluids and metals. The ca. 120 Ma

30 Au mineralization has been previously proposed to have occurred during a

31 magmatic lull, which would negate a magmatic-hydrothermal genetic model.

32 However, recent drilling has revealed a buried mineralized monzonite

33 equivalent in age to the Au mineralization, in the Linglong goldfield. Here, we

34 present comprehensive textural, geochemical (LA-(MC)-ICPMS trace element,

35 Nd and S isotopes) and geochronological (LA-ICPMS U-Pb dating) analyses of

36 titanite and pyrite from this previously unrecognized monzonite. Three types of

37 titanite were distinguished, including magmatic Ttn1 and hydrothermal Ttn2

38 and Ttn3, which show indistinguishable U-Pb ages (120.7 ± 3.1 Ma and 120.9

39 ± 2.6 Ma), REE patterns and Nd isotopes (εNd(t) = -14.7 to -12.9), implying that

40 hydrothermal fluids were directly exsolved from the monzonitic magma,

41 contemporaneous with the large-scale Au mineralization at ca. 120 Ma. The Nd

42 isotopes of titanite potentially indicate a lower crustal source mixed with mantle

43 materials for the monzonite. Four types of pyrite were analyzed, including

44 magmatic Py1 from the fresh biotite monzonite, hydrothermal Py2 from the

45 altered biotite monzonite, hydrothermal Py3 from quartz-pyrite veins with a

46 monazite U-Pb age of 118.2 ± 4.6 Ma, and magmatic Py4 from mafic enclaves

47 of the Gushan granite at ca.120 Ma. The d34S values of magmatic Py1 and Py4

48 (+1.9 to +6.3 ‰, and +5.0 to +6.4 ‰, respectively) and hydrothermal Py2 and

49 Py3 (+6.4 to +9.5 ‰, and +6.5 to +7.6 ‰, respectively) are consistent with

50 sulfur isotopic fractionation between melt and fluid. Hydrothermal Py2 and Py3

51 also have higher Co, As, Ag, Sb and Bi contents and submicron gold inclusions,

52 implying that the magmatic-hydrothermal fluids were fertile for mineralization.

53 This study highlights the importance of monzonite magmatism and exsolved

54 fertile fluids in regional Au mineralization. Hydrous magmas at ca.120 Ma

55 probably extracted Au efficiently from the lower crustal-mantle sources and

56 released auriferous fluids at the late magmatic stage, leading to the formation

57 of Au deposits in the Jiaodong province.

58 Keywords: Titanite, pyrite, monazite, biotite monzonite, U-Pb geochronology,

59 magmatic-hydrothermal fluid

61 INTRODUCTION

62 The Early Cretaceous Au deposits in the Jiaodong Peninsula constitute the

63 largest Au province of China, containing more than 5000 tons of proven Au

64 resources. Mineralization in the Jiaodong Peninsula is considered unique

65 relative to the accepted model for orogenic Au deposits (e.g., Groves et al.,

66 1998), in that Au mineralization postdates the Precambrian metamorphic

67 wallrocks by about 2 billion years (Goldfarb and Santosh 2014). Numerous

68 studies have defined the ore-controlling structures, hydrothermal alterations,

69 ore-fluid chemistry, hydrothermal processes and fluid evolution of both

70 disseminated- to stockwork-style and auriferous quartz-pyrite vein-type Au

71 deposits across the Jiaodong Peninsula (Fan et al. 2003, 2007; Li et al. 2012a;

72 Goldfarb and Santosh 2014; Mills et al. 2015; Deng et al. 2015, 2020b; Feng et

73 al. 2018, 2020; Li et al. 2018). The timing of Au mineralization has also been

74 well constrained to 120 ± 5 Ma by Rb-Sr dating of pyrite, 40Ar/39Ar dating of

75 hydrothermal muscovite and U-Pb dating of hydrothermal monazite (Yang and

76 Zhou 2001; Ma et al. 2017; Li et al. 2018; Zhang et al. 2020; Deng et al. 2020a).

77 Nevertheless, the genesis of the Jiaodong Au province is contentious, given the

78 failure of current models of mineralization to account for its features. This

79 debate is primarily rooted in uncertainties as to the sources of ore-forming fluids

80 and metals, and particularly their genetic relationship(s) to Mesozoic

81 magmatism, Precambrian metamorphic host rocks, subducted slab and

82 sediment, and/or metasomatized lithosphere (Zhai et al. 2004; Chen et al. 2005;

83 Goldfarb and Santosh 2014; Zhu et al. 2015; Deng et al. 2015, 2020b; Wang et

84 al. 2020).

85 Two main problems obstruct linking Au mineralization to the widespread

86 granitic magmatism in the Jiaodong Au province. Firstly, a time gap of at least

87 ~8 million years separates the ages of mineralization and of known magmatism

88 (Li et al. 2019a; Deng et al. 2020a). Secondly, direct evidence for a genetic

89 affinity between magmatism and hydrothermal mineralization has been hitherto

90 lacking. However, recent deep drilling projects in the Linglong goldfield, one

91 of the largest Au-producing districts in the Jiaodong Au province, discovered a

92 buried mineralized biotite monzonite that shows petrographic features distinct

93 from those of the exposed Mesozoic granitoids (Linglong, Luanjiahe,

94 Guojialing and Aishan granites). Shen et al. (2016) constrained the

95 emplacement age of the monzonite to ca. 124 Ma by 40Ar/36Ar dating of biotite,

96 and suggested a temporal relationship between the monzonite and Au

97 mineralization. The monzonite is nearly contemporaneous with the regional

98 Gushan granite, which has a zircon U-Pb age of 120-119 Ma (Li et al. 2012b).

99 Thus, this newly discovered monzonite suggests a potentially extensive episode

100 of magmatism during this period. Moreover, the mineralized monzonite shows

101 distinctive hydrothermal features, suggesting a relationship between magmatic-

102 hydrothermal fluids and mineralization. As such, the mineralized monzonite

103 provides a good opportunity to evaluate the role of magmatism and magmatic-

104 hydrothermal fluid release in the ore-forming process, and to explore the

105 sources of ore-forming fluids and metals, in the Jiaodong Au province.

106 Several magmatic-hydrothermal minerals are present in the monzonite,

107 including titanite, monazite, pyrite, pyrrhotite and magnetite. Titanite

108 commonly contains sufficient uranium contents for precise U-Pb dating and has

109 been widely used to date magmatic and hydrothermal events (Frost et al. 2001;

110 Li et al. 2010; Xiao et al. 2021). Incorporation of minor and trace elements, such

111 as V, M n, Cu, Zr, REE, Hf, Pb, Th and U, in titanite is sensitive to temperature,

112 pressure, oxygen fugacity, and melt and fluid compositions. As such, titanite

113 can be a valuable indicator of magmatic and hydrothermal conditions (Hayden

114 et al. 2008; Xie et al. 2010; Cao et al. 2015; Xiao et al. 2021). The trace elements

115 and sulfur isotopes of pyrite and pyrrhotite can be used to discriminate different

116 sulfur sources and genetic environments, and to understand the

117 physicochemical conditions of sulfide deposition (Deditius et al. 2014; Peterson

118 and Mavrogenes 2014; Tanner et al. 2016; Li et al. 2018). Similarly, magnetite

119 is stable over a wide range of magmatic-hydrothermal conditions and has been

120 used to derive insights on petrogenetic discrimination, physicochemical

121 conditions, and the magma fertility (Dare et al. 2014; Zhi et al. 2019; Dora et

122 al. 2020). The geochemistry of these magmatic-hydrothermal minerals, when

123 integrated with detailed microtextural studies, is useful in fingerprinting

124 magmatic and hydrothermal processes.

125 Here, we document results of detailed textural studies, in situ U-Pb

126 geochronology and trace-element and isotope geochemistry of magmatic-

127 hydrothermal titanite, monazite, pyrite, pyrrhotite and magnetite from the

128 mineralized biotite monzonite, as well as pyrite in mafic enclaves from the

129 Gushan pluton. Our results provide insight into: (1) the temporal relationship

130 between the monzonite and Au mineralization; (2) the link between magmatic

131 and hydrothermal events; (3) the mineralization potential of the monzonite; and

132 (4) the sources of the ore-forming fluids and metals in the Jiaodong Au province.

133 GEOLOGIC BACKGROUND

134 Regional geology

135 The Jiaodong Peninsula is located at the southeastern margin of the North

136 China Craton (NCC), bounded by the crustal-scale Tancheng-Lujiang (Tan-Lu)

137 fault to the west. The peninsula comprises the Jiaobei Terrane in the northwest

138 and the Sulu ultrahigh-pressure metamorphic (UHP) belt in the southeast, which

139 are separated by the Wulian-Yantai Fault (Zhou et al. 2008; Fig. 1). The

140 northern Jiaobei Uplift on the Jiaobei Terrane is dominated by Precambrian

141 metamorphic basement, including the Neoarchean Jiaodong Group, the

142 Paleoproterozoic Jingshan and Fenzishan Groups, and the Neoproterozoic

143 Penglai Group. These rocks were intruded by widespread middle-late Mesozoic

144 granitoids (the Jurassic Linglong and Luanjiahe granite at 160-159 Ma, the

145 Cretaceous Guojialing granodiorite at 130-126 Ma, Gushan granite at 120-119

146 Ma and Aishan-Sanfoshan granite at 118-115 Ma; Miao et al. 1997; Goss et al.

147 2010; Li et al. 2012b; Yang et al. 2012; Li et al. 2019a; Fig. 1), whereas the

148 southern Jiaolai Basin is filled with Cretaceous sedimentary and volcanic rocks

149 (Liu et al. 2009; Xie et al. 2012). The Sulu UHP belt mainly consists of

150 Neoproterozoic granitic gneisses, Triassic UHP metamorphic rocks and

151 Jurassic-Cretaceous granitoids (Zheng 2008; Xu et al. 2016). Structurally, the

152 Jiaodong Peninsula is dominated by NE- to NNE-trending brittle-ductile shear

153 zones with sinistral-oblique-reverse movements in the Late Jurassic, followed

154 by late reactivation and development of brittle normal faults in the Early

155 Cretaceous (Sun et al. 2007; Deng et al. 2015).

156 Deposits in the Jiaodong Au province are divided into three belts from west

157 to east: the Zhaoyuan-Laizhou, Penglai-Qixia and Muping-Rushan belts. The

158 deposits are distributed along regional second- or third- order NE- to NNE-

159 trending faults of the Tan-Lu fault system, particularly the Sanshandao-

160 Cangshang, Jiaojia-Xincheng, Zhaoyuan-Pingdu, Qixia and Muping-Rushan

161 faults (Fig. 1; Fan et al. 2007). The majority of Au deposits, dated at 120 ± 5

162 Ma, are hosted within the widespread Jurassic Linglong, Luanjiahe and

163 Kunyushan granites, Cretaceous Guojialing granodiorite and Precambrian

164 metamorphic rocks (e.g., Yang and Zhou 2001; Fan et al. 2003; Li et al. 2012a,

165 2018; Zhang et al. 2020; Deng et al. 2020a). Quartz-sulfide vein-type and

166 disseminated/stockwork-type Au mineralization account for most of the Au

167 resources in the Jiaodong Au province. These deposits are thought to have

168 formed within the same tectonic setting, but under different local stress fields

169 (Qiu et al. 2002; Yang et al. 2018). Of the three belts, the Zhaoyuan-Laizhou

170 belt is the most significant in terms of endowment, hosting over 85% of Au

171 resources in the district.

172 Ore geology and petrography

173 The Linglong goldfield, which contains several large Au deposits such as

174 Linglong, Jiuqu and Dongfeng, is located to the east of the Zhaoyuan-Laizhou

175 Au belt and at the northern tip of the Zhaoyuan-Pingdu fault (Fig. 1). This

176 goldfield is typified by the quartz-sulfide vein-type mineralization, with minor

177 disseminated sulfide replacements and stockworks. Hundreds of auriferous

178 quartz veins are present in this goldfield, occurring in steeply dipping, NE- to

179 NNE-trending fractures in the hanging wall of the Zhaoyuan-Pingdu fault. The

180 mineralized veins are mainly hosted in the Jurassic Linglong and Luanjiahe

181 granites, with fewer in the Cretaceous Guojialing granodiorite (Wen et al. 2015).

182 The Zhaoyuan-Pingdu fault is intersected by the NEE-trending Potouqing fault

183 and crosscut by the NNE-trending Linglong fault, both of which control the

184 occurrence of ore bodies and the widespread emplacement of intermediate to

185 mafic dykes.

186 The mineralized biotite monzonite was found in drill cores 72ZK1 and

187 72ZK2, at depths of 2070 m and 2374 m separately in the Linglong goldfield,

188 and was suggested as a buried intrusion into the Jurassic Linglong and

189 Luanjiahe granite (Fig. 2; Shen et al. 2016). The biotite monzonite is dark gray

190 to light grayish green, massive, and medium-to-fine grained. Pyrite, pyrrhotite

191 and magnetite are evenly distributed in the monzonite, with some grains

192 occurring as clusters (Fig. 3a-d, g). The mineral assemblage of the monzonite

193 mainly comprises K-feldspar (35-40%), plagioclase (20-25%) and biotite (15-

194 20%) (Fig. 3g, h). Other minerals include quartz (5-10%), amphibole (<5%),

195 titanite, and minor zircon, allanite, apatite. Some coarse grains of K-feldspar

196 and plagioclase have been altered to clay minerals, sericite and chlorite (Fig.

197 3h).

198 Silicification, sericitization and carbonation occurred in the biotite

199 monzonite near the intrusive contact with the overlying Jurassic Luanjiahe

200 granite, whereas unaltered rocks were present at greater depth. The alteration is

201 characterized mostly by addition of quartz, sericite, chlorite and calcite. Several

202 quartz-pyrite veinlets are observed in the overlying Luanjiahe granite (Fig. 3e),

203 suggesting a hydrothermal event probably related to the emplacement and

204 degassing of the monzonitic magma.

205 The Gushan granite, located in the southwest of the Linglong goldfield,

206 approximately 35 km away from the drill core 72ZK1 (Fig. 1), intrudes the

207 Jurassic Luanjiahe granite and carries mafic microgranular enclaves (MMEs,

208 Fig. 3f). The MMEs are generally fine grained and contain dominantly

209 hornblende, plagioclase, biotite, K-feldspar and quartz, with minor apatite and

210 titanite. Some pyrite crystals, coexisting with magnetite and chalcopyrite, are

211 found in the MMEs.

212 SAMPLES AND ANALYTICAL METHODS

213 Representative samples were collected from the drill core 72ZK1 in the

214 Linglong goldfield, including quartz-pyrite veinlets (19ZY09), silicified-

215 sericitized-carbonated monzonite (19ZY01, 19ZY07, 19ZY13, 19ZY15) and

216 fresh to weakly-altered biotite monzonite (19ZY03, 19ZY14) at depths 2056-

217 2153 m. Samples of MMEs from the Gushan pluton were also analyzed.

218 Petrographic studies were carried out on thin sections using field emission

219 scanning electron microscope (FESEM). Titanite and hydrothermal monazite

220 were analyzed for geochronology, major and trace element contents, and Nd

221 isotopes using EPMA, LA-ICPMS and LA-MC-ICPMS. Pyrite, pyrrhotite and

222 magnetite were selected for in situ trace elemental and/or sulfur isotopic

223 analyses using LA-ICPMS and LA-MC-ICPMS. Except for FESEM and EPMA

224 analyses, which were done at the Institute of Geology and Geophysics, Chinese

225 Academy of Sciences (IGGCAS), other measurements were all conducted at

226 the Wuhan Sample Solution Analytical Technology Co., Ltd., China.

227 In situ titanite U-Pb dating and geochemical analyses

228 The major element compositions of titanite were determined using a JEOL-

229 JXA8100 electron microprobe. The operating conditions were 15 kV

230 accelerating voltage, 10 nA beam current and 5 μm probe beam. Calibration

231 standards used were diopside for Ca and Si, jadeite for Al and Na, garnet for Fe,

232 bustamite for Mn, K-feldspar for K, tugtupite for Cl, and fluorite for F. The data

233 were corrected using the atomic number-absorption-fluorescence (ZAF)

234 method.

235 Prior to analytical works, titanite grains were carefully examined using

236 transmitted and reflected light microscopy and backscatter electron (BSE)

237 images so as to avoid fractures, inclusions and U-rich domains. In addition,

238 magmatic and hydrothermal titanites were distinguished preliminarily based on

239 their occurrence, texture (oscillatory versus core-rim and irregular zoning) and

240 mineral association. In situ U-Pb isotopes and trace elements of titanite were

241 analyzed simultaneously by LA-ICPMS employing an Agilent 7900 ICP-MS

242 with a Geolas HD laser ablation system. A spot size of 44 μm was used with

243 pulse rate of 3 Hz and laser energy of 80 mJ. The Pb/U ratios were calibrated

244 against the zircon standard 91500 (Wiedenbeck et al. 1995) and monitored

245 according to the titanite standard MKED1 (Spandler et al. 2016). NIST 610 was

246 used as an external standard, and average CaO concentrations of titanite

247 determined by EPMA were used as the internal standard. Each analysis

248 incorporated a background acquisition of approximately 20 s followed by 50 s

249 of data acquisition. Details of the operating conditions, analytical method and

250 data reduction process can be found in Liu et al. (2008). A Microsoft Excel-

251 based spreadsheet ICPMSDataCal was used to perform off-line selection and

252 integration of background and analyzed signal, time-drift correction and

253 quantitative calibration for U-Pb dating and trace element analysis (Liu et al.

254 2008). Tera-Wasserburg Concordia diagrams and weighted mean calculations

255 were made using Isoplot/Ex_ver3 (Ludwig 2003).

256 In situ Nd isotopes of titanite were measured on spots previously analyzed

257 for U-Pb dating using a Neptune Plus multi-collector ICPMS, coupled with a

258 Geolas HD exciter ArF laser ablation system. A laser pulse energy of ~ 90 mJ,

259 repetition rate of 8 Hz and spot size of 32 μm were used. The mass

260 discrimination factor for 143Nd/144Nd was determined using 146Nd/144Nd (0.7219)

261 with an exponential fractionation law. The 149Sm signal was used to correct the

262 remaining 144Sm interference on 144Nd, using the 144Sm/149Sm ratio of 0.2301.

263 The mass fractionation of 144Sm/149Sm was corrected by the 147Sm/149Sm

264 normalization, using the interference-free 147Sm/149Sm = 1.08680 and

265 exponential law (Xu et al. 2015). The εNd(t) values were calculated relative to

266 the chondritic uniform reservoir (CHUR) parameters of Jacobsen and

267 Wasserburg (1980) (present-day 143Nd/144Nd = 0.512638 and 147Sm/144Nd =

268 0.1967) with 143Nd/144Nd renormalized to 146Nd/144Nd = 0.7219 (Hamilton et al.

269 1983; Bouvier et al. 2008).

270 In situ monazite U-Pb dating

271 In situ U-Pb dating of monazite was conducted using the same instruments

272 and analytical methods described above for the titanite analyses. The spot size

273 and frequency of the laser for U-Pb dating were set to 16 µm and 2 Hz. Monazite

274 standard 44069 (Aleinikoff et al. 2006) and glass NIST610 were used as

275 external standards for U-Pb dating and trace element calibration, respectively.

276 In situ trace element analyses of pyrite and magnetite

277 Trace element concentrations of pyrite and magnetite were analyzed by LA-

278 ICPMS. Detailed operating conditions and data reduction are similar to

279 description by Liu et al. (2008). An Agilent 7900 ICP-MS instrument equipped

280 with a Geolas HD laser ablation system was used to acquire ion signal

281 intensities. A “wire” signal smoothing device is included in the laser ablation

282 system (Hu et al. 2015). For trace element analyses of pyrite, the spot size and

283 frequency of the laser were set to 32 µm and 5 Hz, respectively. The sulfide

284 reference material of MASS-1 (USGS) was used as an external bracketing

285 standard. The data quality was monitored by analyzing the NIST 610 reference

286 glass and the MASS-1 as unknowns interspersed with the measurements of the

287 samples. For magnetite, the spot size and frequency of the laser were set to 44

288 µm and 5 Hz, respectively. Trace element compositions were calibrated against

289 various reference materials (BHVO-2G, BCR-2G and BIR-1G) without using

290 an internal standard (Liu et al. 2008). Each analysis incorporated a background

291 acquisition of approximately 20-30 s followed by 50 s of data acquisition. An

292 Microsoft Excel-based spreadsheet ICPMSDataCal was used in the identical

293 fashion as indicated for trace element measurements above (Liu et al. 2008).

294 In situ S isotope analyses of pyrite and pyrrhotite

295 In situ sulfur isotope analyses of pyrite and pyrrhotite were performed on a

296 Neptune Plus MC-ICPMS equipped with a Geolas HD excimer ArF laser

297 ablation system. Pyrite and pyrrhotite were both ablated using a large spot size

298 (44 μm) and slow pulse frequency (2 Hz) to avoid the down-hole fractionation

299 effect (Fu et al., 2016). 100 laser pulses were completed in one analysis. The

300 laser fluence was kept constant at ~5 J/cm2. The Neptune Plus was equipped

301 with nine Faraday cups fitted with 1011 Ω resistors. Isotopes 32S, 33S and 34S

302 were collected in Faraday cups using static mode. A standard-sample bracketing

303 method was employed to correct for instrumental mass fractionation. To avoid

304 matrix effects, a pyrite standard PPP-1 (Gilbert et al. 2014) was chosen as a

305 reference material for correcting the natural pyrite and pyrrhotite (Fu et al.

34 306 2016). In addition, the in-house reference pyrite SP-Py-01 (δ SVCDT = 2.0 ‰ ±

34 307 0.5 ‰) and pyrrhotite SP-Po-01 (δ SVCDT = 1.4 ‰ ± 0.4 ‰) were analyzed

308 repeatedly as unknown samples to verify the accuracy of the calibration method.

309 Standard errors for PPP-1, SP-Py-01 and SP-Po-01 are ± 0.2 ‰ (2s; N = 54), ±

310 0.2 ‰ (2s; N = 20) and ± 0.2 ‰ (2s; N = 4), respectively.

311 RESULTS

312 Textures and chemical compositions of titanite

313 Titanite crystals are ubiquitous in the biotite monzonite, showing a wide

314 range of textures. These crystals are about 100-400 μm in length, euhedral to

315 anhedral in shape. Three types of titanite were identified based on their

316 occurrence, petrographic features and chemical compositions (Figs. 4, 5, and

317 Supplementary Fig. A1). The titanite compositional data obtained during this

318 study are provided in Supplementary Table 1.

319 Type 1 titanite (Ttn1) typically displays discrete, euhedral-subhedral and

320 rhombus-shaped grains, and occurs as intergrowths with plagioclase, K-feldspar

321 and biotite in the matrix. Oscillatory zoning is the most common and

322 characteristic internal texture of Ttn1, consisting of alternating light and dark

323 bands observed in BSE images (Fig. 4a, b). In some cases, Ttn1 also shows

324 sector zoning consisting of patchy dark, light and/or oscillatory zones (Fig. 4a).

325 These features suggest that Ttn1 is of magmatic origin. Ttn1 shows limited

326 ranges of CaO (27.7-29.1 wt.%), TiO2 (34.6-37.9 wt.%) and SiO2 (30.2-31.8

327 wt.%), and low contents of Al2O3 (0.94-2.23 wt.%), FeO (0.59-1.57 wt.%) and

328 MnO (0.03-0.24 wt.%) (Fig. 5). Trace element data show that Ttn1 has low

329 concentrations of total REE (SREE = 862-9228 ppm, average 4431 ppm) and

330 HFSE (Nb + Ta + Zr + Hf = 227-4964 ppm, average 1733 ppm), together with

331 slight enrichment of Light REE over Heavy REE (LREE/HREE = 2.8-17.8,

332 average 8.5), low Lu/Hf ratios (0.1-0.6, average 0.3), and weakly negative Eu

333 anomalies (EuN/EuN* = 0.5-0.9, average 0.8) (Fig. 5, Supplementary Fig. A1).

334 Ttn1 shows large variations in the contents of Pb (0.4-12.7 ppm, average 2.2

335 ppm), Th (1.8-398.8 ppm, average 94.0 ppm), U (0.3-518.2 ppm, average 57.6

336 ppm) and Y (60-2420 ppm).

337 Type 2 titanite (Ttn2) occurs as subhedral to anhedral crystals characterized

338 by simple zoning comprising irregular bright cores and light-gray contrasting

339 rims under BSE (Fig. 4c-e). This type of titanite commonly shares planar

340 contacts with pyrite. Ttn2 tends to have lower CaO (24.6-28.6 wt.%), TiO2

341 (31.1-37.1 wt.%) and SiO2 (28.4-30.8 wt.%), but higher FeO (0.83-2.41 wt.%)

342 and MnO (0.13-0.33 wt.%) relative to Ttn1 (Fig. 5). Ttn2 also shows

343 significantly higher concentrations of REE (3221-37648 ppm, average 20766

344 ppm), HFSE (531-5040 ppm, average 2529 ppm), Y (633-4960 ppm, average

345 2643 ppm), Pb (0.7-12.9 ppm, average 5.6 ppm), Th (39.1-738.6 ppm, average

346 376.7 ppm), and U (4.8-291.5 ppm, average 86.0 ppm), but broadly similar

347 enrichment of LREE over HREE with slightly higher ratios of LREE/HREE

348 (11.7-30.3, average 21.4) and Lu/Hf (0.1-2.9, average 0.9) (Fig. 5,

349 Supplementary Fig. A1). In addition, Ttn2 displays more prominent negative

350 Eu anomalies (EuN/EuN* = 0.3-0.7, average 0.50) than Ttn1.

351 Type 3 titanite (Ttn3) occurs as fine, irregular veins penetrating and

352 overgrowing Ttn2 (Fig. 4f-h). Ttn3 commonly occurs together with calcite and

353 displays dark gray color under BSE. Ttn3 has similar ranges of CaO (27.3-28.9

354 wt.%), TiO2 (34.7-37.8 wt.%), SiO2 (30.3-31.3 wt.%) and Al2O3 (0.98-2.05

355 wt.%) to Ttn1, but slightly higher contents of FeO (0.67-2.23 wt.%) and MnO

356 (0.05-0.23 wt.%) (Fig. 5). The contents of trace elements of Ttn3 lie mostly

357 between those of Ttn1 and Ttn2，with REE (SREE = 1173-28457 ppm, average

358 10144 ppm), HFSE (123-8136 ppm, average 2344 ppm), Pb (0.5-6.8 ppm,

359 average 2.8 ppm), Th (14.8-511.2 ppm, average 157.4 ppm), U (1.7-197.0 ppm,

360 average 36.3 ppm) and Y (197-5389 ppm, average 2049 ppm). Ttn3 also

361 displays moderate Lu/Hf ratios (0.2-1.5, average 0.5), and a slight enrichment

362 of LREE relative to HREE (LREE/HREE = 6.2-18.6, average 13.0), showing

363 flat chondrite-normalized REE patterns with negative Eu anomalies (EuN/EuN*

364 = 0.2-1.1, average 0.6), similar to that of Ttn1 and Ttn2 (Supplementary Fig.

365 A1).

366 Geochronology of the biotite monzonite and hydrothermal event

367 Titanite U-Pb ages

368 In situ titanite U-Pb isotopic data are provided in Table 1. Locations of

369 representative points for U-Pb dating are shown in Fig. 4. The uncorrected data

370 were plotted on Tera-Wasserburg Concordia diagrams (Fig. 6a, b), and a

371 regression through these data yielded a lower-intercept age. The y-intercept

372 represents the initial 207Pb/206Pb ratio which can be used to conduct 207Pb-

373 correction to calculate the corrected 206Pb/238U age (Aleinikoff et al. 2002). The

374 magmatic titanites (Ttn1) define a lower-intercept age of 120.7 ± 3.1 Ma (2s, n

375 = 18, MSWD = 1.8) with initial 207Pb/206Pb = 0.8516. With this common Pb

376 composition, a 207Pb-correction was conducted. Ten points yield a weighted

377 average 206Pb/238U age of 118.8 ± 2.2 Ma (MSWD = 1.3) (Fig. 6a), consistent

378 within error with the lower-intercept age. However, eight points exhibit lower

379 U and Pb with higher percentage of common Pb, plotting away from the lower

380 intercept and resulting in greater experimental errors. They cannot yield

381 convincing 206Pb/238U ages based on 207Pb-correction, thus the lower-intercept

382 age (120.7 ± 3.1 Ma) is used as the magmatic titanite age. The hydrothermal

383 titanites (Ttn2 and Ttn3) give a lower-intercept age of 120.9 ± 2.6 Ma (2s, n =

384 19, MSWD = 1.4) with initial 207Pb/206Pb = 0.8530. Fifteen points define a 207Pb-

385 corrected weighted average 206Pb/238U age of 117.9 ± 2.2 Ma (MSWD = 1.1)

386 which is consistent with the lower-intercept age (Fig. 6b). The other four points,

387 plotting away from the lower intercept, were failed to produce convincing

388 206Pb/238U ages, thus the lower-intercept age (120.9 ± 2.6 Ma) is used as the

389 hydrothermal titanite age. It is noteworthy that the 206Pb/238U ages for Ttn2 and

390 Ttn3, as well as cores and rims in Ttn2, are indistinguishable, thus they were

391 regarded as a whole during age analysis.

392 Monazite U-Pb ages

393 The in situ U-Pb data of 19 spots on monazite grains are presented in Table

394 1. All the analyzed monazite grains are associated or intergrown with pyrite in

395 the quartz-pyrite veinlets (Fig. 4i), thus their crystallization ages represent the

396 timing of the hydrothermal (vein-forming) event. The U-Pb data give a lower-

397 intercept age of 119.0 ± 5.8 Ma (2s, n = 19, MSWD = 1.1) on the Tera-

398 Wasserburg Concordia diagram. The initial 207Pb/206Pb value is 0.7373. The

399 207Pb-corrected weighted average 206Pb/238U age is 118.2 ± 4.6 Ma (MSWD =

400 0.2) which is consistent with the lower intercept age (Fig. 6c).

401 Nd isotopic composition of titanite

402 In situ Nd isotopic compositions of titanite are shown in Supplementary

403 Table 1. The analyzed Ttn1, Ttn2 and Ttn3 show restricted and indistinguishable

404 εNd(t = 121 Ma) values of -14.7 to -13.0, -14.7 to -13.5, and -14.3 to -12.9 (Fig.

405 7).

406 Classification, trace-element and sulfur isotopic compositions of pyrite,

407 magnetite and pyrrhotite

408 Based on occurrence, optical characteristics and mineral assemblages, four

409 types of pyrite were recognized. The complete dataset of trace element

410 concentrations and sulfur isotopes is given in Supplementary Table 2.

411 Type 1 pyrite (Py1) occurs in the fresh biotite monzonite. It is characterized

412 by isolated euhedral-anhedral crystals coexisting with magnetite, and contains

413 inclusions of plagioclase, quartz and titanite (Fig. 8a, e). The Py1 contains

414 uniformly low concentrations of Co (3-374 ppm), Ni (3-41 ppm), Cu (0-7.9

415 ppm), Zn (0-4.1 ppm), As (0.2-0.9 ppm), Ag (0-0.05 ppm), Sb (0-0.09 ppm), Au

416 (0-0.03 ppm, with two outliers at 0.1 ppm and 0.14 ppm), Pb (0-8.5 ppm) and

417 Bi (0-6.1 ppm) (Fig. 9). It commonly displays homogeneous textures on BSE

418 images (Fig. 8e), contrasting with heterogeneous compositional textures of Py2

419 and Py3 (Fig. 8f, g). The d34S values of Py1 are relatively low with a large

420 variation from +1.9 to +6.3 ‰ (average +4.7 ‰) (Fig. 10). The coexisting

421 magnetite grains mostly plot within the magmatic magnetite region on the Ti vs.

422 Ni/Cr diagram (Dare et al., 2014) with low Ni/Cr ratios (0.02-0.93)

423 (Supplementary Fig. A2).

424 Type 2 pyrite (Py2) mostly occurs as anhedral clusters in the altered

425 monzonite which shows significant sericitization and carbonation. The Py2 is

426 commonly enclosed within and crosscut by pyrrhotite (Fig. 8b, f), suggesting

427 that Py2 formed earlier than pyrrhotite. It is likely partly replaced by pyrrhotite,

428 resulting in corroded grain boundaries and metasomatic-relict texture (Fig. 8b).

429 Trace element data show that the contents of Co (0.1-4837 ppm), Ni (0-87.6

430 ppm), As (0.3-86.4 ppm), Ag (0-0.4 ppm), Sb (0-4.5 ppm), Au (0-0.09 ppm,

431 with one outlier of 18.7 ppm), Pb (0-218 ppm) and Bi (0-9.9 ppm, with an

432 outlier of 16.8 ppm) in Py2 have large variations and are higher than Py1 (Fig.

433 9). The d34S values of Py2 are +6.4 to +9.5 ‰ (average +8.4 ‰) (Fig. 10), which

434 are higher than Py1 but similar to the d34S values of pyrite from other deposits

435 in the Jiaodong Au province (+6 to +10 ‰; Mao et al. 2008; Feng et al. 2018;

436 Li et al. 2018; Deng et al. 2020b). The pyrrhotite has a d34S range of +6.9 to

437 +8.8 ‰ (average +7.9 ‰) that overlaps with range of Py2.

438 Type 3 pyrite (Py3) occurs as subhedral, rounded crystals in the quartz-

439 pyrite veins (Fig. 8c, g). The Py3 contains low concentrations of Co (0-98.6

440 ppm), Ni (0-116 ppm), Cu (0-0.5 ppm), Zn (0.4-0.9 ppm), Ag (0-0.3 ppm), Sb

441 (0-0.1 ppm), Au (0-0.01 ppm, with an outlier of 0.1 ppm), Pb (0-2.2 ppm) and

442 Bi (0-1.0 ppm, with an outlier of 19.9 ppm), but markedly high contents of As

443 (325-852 ppm) (Fig. 9). The d34S values of Py3 are +6.5 to +7.6 ‰ (average

444 +7.2 ‰), which is entirely with the range of Py2 (Fig. 10).

445 Another type of pyrite (Py4) was sampled from the mafic enclaves of the

446 Gushan granite. It occurs as euhedral to anhedral crystals in the matrix and

447 commonly coexists with magnetite and minor chalcopyrite (Fig. 8d, h). Most of

448 the trace elements, including Co, Cu, Zn, As, Sb, Au, Pb and Bi, show low

449 concentrations (mostly below detection limits), but the contents Ag (0-2.85 ppm)

450 and Ni (227-3407 ppm) are relatively high and vary widely. The Py4 displays a

451 low and narrow range of d34S values from +5.0 to +6.4 ‰ (average +5.9 ‰)

452 (Fig. 10).

453 DISCUSSION

454 Origins of the different generations of titanite

455 Distinct origins of titanite can be discriminated by their textures,

456 paragenetic mineral assemblages and chemical compositions (Cao et al. 2015;

457 Fu et al. 2016). The Ttn1 grains are mostly euhedral and rhombus-shaped, show

458 oscillatory and sector zoning under BSE, and occur as intergrowths with

459 igneous plagioclase, K-feldspar and biotite in the matrix of the biotite

460 monzonite, implying a magmatic origin. In contrast, the Ttn2 and Ttn3 grains

461 are mostly subhedral to anhedral, show core-rim texture and penetrating

462 relationship under BSE, and commonly show planar contacts with hydrothermal

463 pyrite, sericite and calcite, suggesting a hydrothermal origin. This conclusion is

464 further supported by the chemical compositions of the titanite grains. As

465 demonstrated in Fig. 5, Ttn1 shows low Fe, Mn, REE and high Ti, Ca. The Ttn1

466 data fall in relatively restricted portions of these four diagrams, resembling

467 typical magmatic titanite. In contrast, Ttn2 and Ttn3 are much more

468 compositionally diverse, with relatively higher Fe, Mn, REE and lower Ti, Ca

469 contents: all characteristics of hydrothermal titanite (Morad et al. 2009; Fu et al.

470 2016). The high Fe and Mn contents of Ttn2 and Ttn3 are consistent with Fe-

471 rich hydrothermal fluids from which pyrite and pyrrhotite formed. Overall, Fig.

472 5a and 5c show negative correlations between TiO2 and FeO, CaO and REE,

473 respectively, indicating that appreciable amounts of Fe and REE substitute for

474 Ti and Ca in the titanite structure, which is common in hydrothermal titanite

475 (Cao et al. 2015; Li et al. 2020). The dominant substitution mechanisms are

476 likely to be: (Al, Fe)3+ + (F, OH)- = Ti4+ + O2-, in which the substitution of Fe

477 for Ti predominates, and REE3+ + (Al, Fe)3+ = Ca2+ + Ti4+ (Olin and Wolff 2012;

478 Xie et al. 2010).

479 Our results indicate that the concentrations of trace elements, including REE,

480 HFSE, Th, U and Pb, are quite distinct between magmatic (Ttn1) and

481 hydrothermal (Ttn2 and Ttn3) titanites (Fig. 5). These differences in trace

482 elements are strongly affected by the formation conditions, such as T, fO2, fS2

483 and composition of melt or fluid (Tiepolo et al. 2002; Cao et al. 2015; Xiao et

484 al. 2021). Hydrothermal Ttn2 and Ttn3 show higher HFSE concentrations and

485 Lu/Hf ratios than the magmatic Ttn1, similar to hydrothermal titanites from

486 other regions documented by Fu et al. (2016) and Xiao et al. (2021). Enrichment

487 of HFSE in hydrothermal titanite suggests that HFSE can be mobile in

488 hydrothermal fluids during silicification, sericitization and carbonation.

489 However, in contrast to other studies, the hydrothermal titanites (Ttn2 and Ttn3)

490 have higher SREE compared with magmatic titanite (Ttn1). This implies that

491 the hydrothermal fluids were relatively enriched in REE. This interpretation is

492 also consistent with the ubiquitous presence of monazite in the quartz-pyrite

493 veins.

494 Despite the differences in REE contents, the magmatic and hydrothermal

495 titanites show similar chondrite-normalized REE patterns (Supplementary Fig.

496 A1), which suggest that the hydrothermal fluids that precipitated Ttn2 and Ttn3

497 were most likely exsolved from the same monzonitic magma that crystallized

498 Ttn1. This conclusion is further supported by the overlapping crystallization

499 ages of magmatic Ttn1 (120.7 ± 3.1 Ma) and hydrothermal Ttn2 and Ttn3 (120.9

500 ± 2.6 Ma), and by the indistinguishable εNd(t) values of the three titanite types

501 (Fig. 7). The latter also implies the same Nd isotope composition of the

502 monzonitic magma and hydrothermal fluids, further suggesting the co-genetic

503 origin between magma and fluid.

504 Given its high Nd content, magmatic titanite could shed light on the Nd

505 isotope composition of the magma (e.g., Xie et al. 2010; Cao et al., 2015). The

506 homogeneous εNd(t) values (-14.7 to -13.0) of magmatic Ttn1 are higher than

507 the whole-rock εNd(t) values of the lower crustal-sourced Linglong granite (-

508 21.6 to -19.2), Gushan granite (-18.9 to -18.6) and Aishan granite (-17.9 to -

509 16.2) (Yang et al. 2012; Li et al. 2012b; Li et al. 2019a), but within the range of

510 the Guojialing granodiorite (-16.8 to -11.5), which has been interpreted as

511 deriving from a mixture of mantle and lower crustal sources (Yang et al. 2012;

512 Li et al. 2019a). Additionally, early Cretaceous mafic dikes in the Jiaodong area

513 have εNd(t) that overlap with range of Ttn1 (-18.1 to -13.6; Cai et al. 2013) (Fig.

514 7). Taken as a whole, these data support a mixed mantle and lower crustal source

515 for the monzonitic magma (Fig. 11a).

516 Interpretation of pyrite geochemistry

517 Trace elements and S isotopes of pyrite record valuable information about

518 the source and properties (pH, redox, temperature) of the fluids from which the

519 pyrite formed (Pokrovski et al. 2002; Deditius et al. 2014; Peterson and

520 Mavrogenes 2014; Li et al. 2018). Four types of pyrite identified in this study

521 have been characterized for trace elements and S isotopes, and the implications

522 of these results are discussed here.

523 The isolated Py1 grains show features consistent with a magmatic origin,

524 including: disseminated occurrence in fresh monzonite; coexistence with

525 magmatic magnetite, plagioclase, quartz and titanite; homogeneous BSE

526 response; and low concentrations of trace metals (Co, Ni, Cu, Zn, As, Ag, Sb,

527 Pb and Bi). The Py1 grains also show several similarities with Py4 grains found

528 in the fresh mafic enclaves from the Gushan granite. The Py4 occurs as euhedral

529 to anhedral crystals in the matrix and coexists with magnetite and allanite (Fig.

530 8d, h). Both Py1 and Py4 are therefore interpreted as being of magmatic origin.

531 Both Py1 and Py4 also have broadly similar and relatively low values of d34S

532 (+1.9 to +6.3 ‰ and +5.0 to +6.4 ‰, respectively), which represent the sulfur

533 isotopic compositions of the magma. Previous research has suggested that the

534 mafic enclaves from the Gushan granite were formed by mixing between mafic

535 and felsic magmas (Li et al. 2012b), during which both chemical and isotopic

536 compositions tend to reach homogenization. The felsic magma was sourced

537 from the ancient NCC lower crust at ca. 120 Ma, and the mafic magma was

538 derived from an enriched lithospheric mantle source beneath eastern NCC (Li

539 et al. 2012b). Similarly, the above results of Nd isotopes in titanite suggest a

540 potentially mixed mantle/lower crustal source for the monzonitic magma at ca.

541 121 Ma. Thus, we interpret that the sulfur isotopic compositions of Py1 and Py4

542 represent the products of isotopic exchange between lower crust-derived felsic

543 magma and lithospheric mantle-derived mafic magma.

544 In contrast to Py1 and Py4, Py2 and Py3 show features typical of

545 hydrothermal origin. Py2 occurs mostly as anhedral clusters in the altered

546 monzonite, coexists with sericite, calcite, and pyrrhotite, and displays

547 heterogeneous textures under BSE (Fig. 8f). In addition, Py2 shows variable

548 and relatively high concentrations of minor and trace metals (especially Co, As,

549 Ag, Sb and Bi). Significantly, extremely high contents of Au (18.7 ppm) and Bi

550 (16.8 ppm) suggest the presence of gold as submicron inclusions in Py2. The

551 latter implies that the hydrothermal fluids that deposited Py2 had potential for

552 forming precious-metal mineralization. Py3 occurs in the quartz-pyrite veins,

553 and thus is unambiguously of hydrothermal origin. Py3 shows various broad

554 similarities with pyrite from the Jiaodong Au province, especially high As

555 contents and d34S values of +6.5 to +7.6 ‰.

556 Pyrrhotite grains commonly occur as rims on or fracture fill within the Py2

557 grains, implying that pyrrhotite formed later than Py2. However, the pyrrhotite

34 34 34 558 and Py2 display similar d S values (d Spo = +6.9 to +8.8 ‰, d SPy2 = +6.4 to

559 +9.5 ‰), which are indicative of natural sulfur isotopic fractionation between

560 pyrite and pyrrhotite precipitated from hydrothermal fluids with the same sulfur

561 ratios (d34S) (e.g., Li et al. 2017). The precipitation of early pyrite transitioning

562 to later pyrrhotite suggests a decrease of sulfur fugacity in the evolving

563 hydrothermal fluid, probably due to the consumption of sulfur by pyrite, or

564 perhaps by effervescence of H2S.

565 Considering that the hydrothermal sulfides disseminated in the monzonite

566 are dominated by pyrite and pyrrhotite, isotopic fractionation between sulfide

567 pairs is negligible, thus the sulfur isotopic signature of pyrite can reasonably

568 represent the bulk sulfur isotopic composition of the hydrothermal fluid

569 (Ohmoto 1972). Sulfur isotopes of pyrite have been widely used to trace the

570 sources of fluids (e.g., Peterson and Mavrogenes 2014; Chen et al. 2015). Based

571 on the sulfur isotopic ratios of Py2 and Py3, we suggest that the corresponding

572 hydrothermal fluids have a narrow range of d34S values (+6.4 to +9.5 ‰),

573 similar to those from the Au deposits in the Jiaodong province (+6 to +10 ‰;

574 Li et al. 2018; Feng et al. 2018; Deng et al. 2020b), but slightly heavier than the

575 monzonitic magma which was represented by the d34S of Py1 (+1.9 to +6.3 ‰)

576 (Fig. 10). The textural, geochemical and geochronological observations of the

577 magmatic-hydrothermal titanite indicate a role of magmatic-hydrothermal

578 fluids that exsolved from the monzonitic magma. These magmatic-

579 hydrothermal fluids are likely to have contributed to the precipitation of Py2

580 and Py3, as well as to silicification, sericitization and carbonation of the biotite

581 monzonite. Experimental data from Fiege et al. (2014) revealed that the

582 hydrothermal fluids released by magma degassing can have heavier S isotopes

583 than the source magma (with d34S ~6.7‰ higher) when the sulfur occurs as S2-

34 584 or H2S. This is compatible with the slightly heavier d S values of the

585 hydrothermal Py2 and Py3 compared to the magmatic Py1. In addition, the age

586 of hydrothermal monazite (118.2 ± 4.6 Ma) is consistent with the age of biotite

587 monzonite (120.7 ± 3.1 Ma), further implying an intimate relationship between

588 the monzonitic magma and hydrothermal fluids. In general, we propose that the

589 hydrothermal fluids that precipitated Py2 and Py3 were directly exsolved from

590 the monzonitic magma, which had previously crystallized igneous Py1 grains.

591 Constraints on magmatism and hydrothermal processes

592 The deposition of titanite and pyrite in the biotite monzonite clearly spans

593 the magmatic and hydrothermal stages (Fig. 11). Our results regarding the

594 occurrence and geochemistry of titanite and pyrite provide insights into the

595 melt/fluid composition, the conditions under which they formed, and the

596 magmatic-hydrothermal processes in the monzonite system.

597 Magmatic Ttn1 formed as discrete euhedral grains intergrown with K-

598 feldspar, magnetite and quartz. This assemblage indicates that the monzonitic

599 magma crystallization occurred at a relatively high oxygen fugacity and H2O-

600 rich environment (Frost and Lindsley 1992; Piccoli et al. 2000; Xiao et al. 2021).

601 The high oxygen fugacity is supported by the negative Eu anomaly of Ttn1.

602 High oxygen fugacity and H2O contents could lead to a preferential transfer of

603 Au into the magma from its source (Botcharnikov et al. 2011; Wang et al. 2020),

604 resulting in Au enrichment, which is reflected by two analyses of magmatic Py1

605 that showed relatively high Au contents (0.1 ppm and 0.14 ppm). Since the Nd

606 isotopes of Ttn1 and S isotopes of Py1 are proposed to be likely the result of

607 mixing between magmas from lower crust and lithospheric mantle, the Au could

608 be derived from both sources (Fig. 11a).

609 In the later hydrothermal stages, fluids were directly exsolved from the late-

610 stage monzonitic magma. The complex textures and geochemistry of Ttn2 and

611 Ttn3 suggest that formation of these minerals was a dynamic process involving

612 different hydrothermal stages (Fig. 11b). The initial hydrothermal fluid was

613 characterized by high REE contents (especially LREE), which resulted in back-

614 scatter bright cores of Ttn2. With further degassing, the hydrothermal fluid

615 precipitated back-scatter gray rims of Ttn2 with moderate REE contents (e.g.,

616 Fu et al. 2016; Xiao et al. 2021), as well as Py2 with relatively high metal

617 contents (Co, As, Ag, Au, Sb and Bi). This implies that Au, together with Co,

618 As, Ag, Sb and Bi, were pre-enriched in the monzonitic magma and were

619 furthermore favorably partitioned into the exsolved fluids (Pokrovski et al.

620 2013). In the late hydrothermal stage, the fluid became less oxidized –

621 evidenced by the variable negative-positive Eu anomalies (EuN/EuN* = 0.2-1.1,

622 average 0.6) of Ttn3 – and contained lower total REE contents. The Ttn3

623 precipitated at this stage overgrew on the rims of Ttn2, and partly penetrated in

624 Ttn2 (Fig. 11b). The pyrrhotite grains were likely precipitated during the late

625 hydrothermal stages (Fig. 11b), due to the consumption of sulfur by deposition

626 of Py2 and less oxidizing conditions. Multiple stages of hydrothermal fluid

627 exsolution from the monzonitic magma show the same Nd and S isotopic

628 compositions. The hydrothermal fluids led to distinct alteration in the biotite

629 monzonite, including silicification, sericitization and carbonation. We surmise

630 that these same fluids then migrated along faults and fractures, forming quartz-

631 pyrite (Py3) veins in the overlying Luanjiahe granite.

632 A potential magmatic-hydrothermal model for the Jiaodong Au

633 mineralization

634 The source(s) of ore-forming fluids and Au, the mechanism(s) for Au

635 enrichment, and the genetic model for the Jiaodong Au deposits have long been

636 debated (Fan et al. 2003; Zhai et al. 2004; Chen et al. 2005; Goldfarb and

637 Santosh 2014; Zhu et al. 2015; Deng et al. 2020b). The possible sources for the

638 ore components include (1) a crustal source from Precambrian metamorphic

639 rocks and/or Mesozoic granite (Chen et al. 2005), (2) a subcrustal source from

640 volatile- and Au-fertilized metasomatized mantle lithosphere (Zhu et al. 2015;

641 Deng et al. 2020a, b), or (3) a mixed crustal plus mantle source (Zhai et al.

642 2004). Various models have been postulated to explain the nature of Au

643 mineralization in different settings, including the orogenic Au deposits (Zhou

644 et al. 2002; Goldfarb and Santosh 2014), intrusion-related Au deposits (Sillitoe

645 and Thompson 1998; Nie et al. 2004), magmatic-hydrothermal Au deposits

646 related to mafic dikes (Wang et al. 2020), decratonic Au deposits (Zhu et al.

647 2015), and “Jiaodong-type” Au deposits (Deng et al. 2015; Li et al. 2015). All

648 these genetic models have advantages and problems, which have been reviewed

649 by Deng et al. (2020b), and are still highly contentious. One critical factor

650 hindering universal acceptance of a single model is the uncertainty of genetic

651 relationship between magmatism and Au mineralization. Ultimately, this

652 uncertainty is underpinned by the question of whether exsolution and degassing

653 of auriferous magmatic-hydrothermal fluid took place. The information

654 provided by magmatic-hydrothermal titanites and pyrites in this study

655 highlights the exsolution of auriferous hydrothermal fluids from the monzonitic

656 magma, and thus supports a magmatic-hydrothermal model for the genesis of

657 the Jiaodong Au province, as discussed below.

658 Titanite and monazite that formed from the monzonitic magma (Ttn1) and

659 exsolved hydrothermal fluids (Ttn2, Ttn3 and monazite) yielded consistent U-

660 Pb ages (121-118 Ma) overlapping within error with the large-scale Au

661 mineralization age at Jiaodong (120-119 Ma; Ma et al. 2017; Li et al. 2018;

662 Yang et al. 2018; Deng et al. 2020a; Zhang et al. 2020). Spatially, the biotite

663 monzonite intruded at depth beneath the Linglong goldfield, and is connected

664 with the surface and Au deposits by the Zhaoyuan-Pingdu fault which hosts

665 hundreds of auriferous quartz veins (Fig. 2b). Hence, our results suggest that

666 the monzonitic magmatism and its accompanying hydrothermal fluids were

667 both temporally and spatially linked to the Au mineralization. Magmatic pyrite

668 (Py1) from the monzonite has relatively high Au contents (up to 0.14 ppm),

669 suggesting that the magma was fertile for Au. Hydrothermal fluids precipitating

670 Py2 have higher Au contents, as reflected by the presence of submicron

671 inclusions of Au and Bi in Py2, again corroborating a genetic relationship with

672 the Au mineralization. Quantitatively, the biotite monzonite may be part of a

673 large-scale magmatic event at ca. 121 Ma in the Jiaodong Au province,

674 consistent with contemporaneous intermediate-felsic rocks, including the

675 Gushan granite located ~35 km to the southwest of the drill core (120-119 Ma;

676 Li et al. 2012b), the widespread intermediate-mafic (130-120 Ma; Ma et al.

677 2017; Wang et al. 2020) and felsic dikes (123-118 Ma; Li et al. 2019b).

678 Extensive magmatism has the potential to produce hydrothermal fluids forming

679 giant Au deposits. Although many researchers proposed that the ore-forming

680 fluids were sourced from the volatile-enriched metasomatized mantle

681 lithosphere (e.g., Goldfarb and Santosh 2014; Deng et al. 2015, 2020a, b),

682 analyses of mantle xenoliths have suggested relatively low Au contents of the

683 metasomatized mantle lithosphere (Wang et al., 2020), casting doubt on that

684 interpretation. Wang et al. (2020) underscored the high Au contents of hydrous

685 magmas that were sourced from the metasomatized mantle. These findings

686 consistently reveal a genetic relationship between biotite monzonite, which

687 potentially involves mantle components, and Au mineralization in the Jiaodong

688 Au province.

689 Previous studies have established that the Jiaodong Peninsula experienced

690 a transition of tectonic regime from regional compression to transpression or

691 transtension prior to peak extension since 125-120 Ma, as a far-field response

692 to the change of subduction direction and roll-back of the paleo-Pacific Plate

693 (Sun et al. 2007; Yang et al. 2018; Deng et al. 2020b). The mantle lithosphere

694 under the eastern NCC at this stage had been metasomatized during Triassic and

695 Late Jurassic to Early Cretaceous subduction events (Chen et al. 2005; Liang et

696 al. 2019; Deng et al. 2020a, b). The extension-induced thinning of lithosphere

697 and accompanying asthenospheric upwelling at 130-120 Ma (Zhai et al. 2004;

698 Li et al. 2019a; Wu et al. 2019) triggered partial melting of the metasomatized

699 mantle lithosphere and the lower crust of the NCC, producing hydrous mafic

700 and felsic magmas, respectively (Fig. 11a). These magmas probably extracted

701 Au efficiently from their sources (Wang et al. 2020). Distinct degrees of mixing

702 occurred between the mafic and felsic magmas, forming the biotite monzonite

703 and Gushan granite (Fig. 11a), with distinct ranges of εNd(t), which likely

704 represent two different intrusives formed by the same magmatism. Attributing

705 to the more involved mafic magmas, higher εNd(t) values were yielded in the

706 biotite monzonite than the Gushan granite. With further magmatic evolution,

707 hydrothermal fluids were exsolved from these crystallizing magmas (Fig. 11a),

708 preferentially concentrating Au, As, Sb, Bi, S, Cl and C into the fluids by a

709 factor of hundreds (Pokrovski et al. 2013). The auriferous fluids ascended along

710 the translithospheric faults into second-order fault systems, and precipitated

711 gold, quartz and sulfides. This model favors a magmatic-hydrothermal origin,

712 and explains the intimate spatial and temporal association of Au deposits with

713 mafic-felsic magmatism.

714 IMPLICATIONS

715 The biotite monzonite buried in the Linglong goldfield formed at ca. 121

716 Ma based on titanite U-Pb dating, and displays distinctive petrographic features

717 indicative of hydrothermal alteration and mineralization, containing ubiquitous

718 magmatic-hydrothermal titanite and pyrite. Three types of titanite with different

719 textures and geochemistry were identified, including magmatic Ttn1 with

720 oscillatory and sector zoning under BSE, and hydrothermal Ttn2 and Ttn3

721 which show core-rim or irregular textures, and higher Fe, Mn, REE, HFSE, but

722 lower Ti, Ca contents than Ttn1. Hydrothermal fluids that precipitated Ttn2 and

723 Ttn3 were sourced from the monzonitic melt that crystallized Ttn1, evidenced

724 by similar chondrite-normalized REE patterns, indistinguishable Nd isotopes

725 (εNd(t) = -14.7 to -12.9) and U-Pb ages (120.7 ± 3.1 Ma for Ttn1, 120.9 ± 2.6

726 Ma for Ttn2 and Ttn3).

727 Four types of pyrite were analyzed, including magmatic Py1 from the fresh

728 biotite monzonite, hydrothermal Py2 from the altered biotite monzonite,

729 hydrothermal Py3 from the quartz-pyrite veins, and magmatic Py4 from the

730 mafic enclaves of the Gushan granite. The Py1 and Py4 have low concentrations

731 of trace metals, while the Py2 and Py3 have high Co, As, Ag, Au, Sb and Bi

732 contents which corresponds to fertile hydrothermal fluids. Magmatic Py1 has

733 broadly similar d34S values (+1.9 to +6.3 ‰) to Py4 (+5.0 to +6.4 ‰), but

734 relatively lighter than hydrothermal Py2 and Py3 (+6.4 to +9.5 ‰), consistent

735 with sulfur isotopic fractionation between melt and fluid. U-Pb dating of

736 monazite (118.2 ± 4.6 Ma) from the quartz-pyrite veins indicate that the

737 auriferous hydrothermal fluids are synchronous to the monzonitic magma.

738 In this contribution, evidences from the titanite and pyrite corroborate the

739 exsolution of fertile hydrothermal fluids from the monzonitic magma,

740 suggesting a magmatic-hydrothermal genetic model for the giant Jiaodong Au

741 mineralization. We infer that primary Au was efficiently extracted from the

742 metasomatized mantle lithosphere and the lower crust of the NCC by hydrous

743 magmas at ca. 120 Ma.

744 ACKNOWLEGMENTS AND FUNDING

745 We thank Dapeng Li and Guijun Wen for their helps during fieldworks,

746 Lihui Jia, Jiakang Kong and Zheng Liu for their technical supports in running

747 the EPMA, LA-ICPMS and LA-MC-ICPMS analyses. We are grateful to two

748 anonymous reviewers for their constructive feedback that helped to improve an

749 earlier manuscript, and associate editor Paul Tomascak for valuable suggestions

750 and efficient editorial handling. This work is funded by the National Key

751 Research and Development Program (No. 2016YFC0600105), China National

752 Postdoctoral Program for Innovative Talents (No. BX20190327), and China

753 Postdoctoral Science Foundation (No. 2019M660788).

754

755 REFERENCES CITED

756 Aleinikoff, J.N., Wintsch, R.P., Fanning, C.M., and Dorais, M.J. (2002) U-Pb

757 geochronology of zircon and polygenetic titanite from the Glastonbury

758 Complex, Connecticut, USA: an integrated SEM, EMPA, TIMS, and

759 SHRIMP study. Chemical Geology, 188, 125–147.

760 Aleinikoff, J.N., Schenck, W.S., Plank, M.O., Srogi, L., Fanning, C.M., Kamo,

761 S.L., and Bosbyshell, H. (2006) Deciphering igneous and metamorphic

762 events in highgrade rocks of the Wilmington Complex, Delaware:

763 Morphology, cathodoluminescence and backscattered electron zoning, and

764 SHRIMP U-Pb geochronology of zircon and monazite. Geological Society

765 of America Bulletin, 118, 39–64.

766 Botcharnikov, R.E., Linnen, R.L., Wilke, M., Holtz, F., Jugo, P.J., and Berndt,

767 J. (2011) High gold concentrations in sulphide-bearing magma under

768 oxidizing conditions. Nature Geoscience, 4, 112–115.

769 Bouvier, A., Vervoort, J.D., and Patchett, P.J. (2008) The Lu–Hf and Sm–Nd

770 isotopic composition of CHUR: Constraints from unequilibrated

771 chondrites and implications for the bulk composition of terrestrial planets.

772 Earth and Planetary Science Letters, 273, 48–57.

773 Cai, Y.C., Fan, H.R., Santosh, M., Liu, X., Hu, F.F., Yang, K.F., Lan, T.G., Yang,

774 Y.H., and Liu, Y.S. (2013) Evolution of the lithospheric mantle beneath the

775 southeastern North China Craton: constraints from mafic dikes in the

776 Jiaobei terrain. Gondwana Research, 24, 601–621.

777 Cao, M.J., Qin, K.Z., Li, G.M., Evans, N.J., and Jin, L.Y. (2015) In situ LA-

778 (MC)-ICP- MS trace element and Nd isotopic compositions and genesis of

779 polygenetic titanite from the Baogutu reduced porphyry Cu deposit,

780 Western Junggar, NW China. Ore Geology Reviews, 65, 940–954.

781 Chen, L., Li, X.H., Li, J.W., Hofstra, A.H., Liu, Y., and Koenig, A.E. (2015)

782 Extreme variation of sulfur isotopic compositions in pyrite from the

783 Qiuling sediment-hosted gold deposit, West Qinling orogen, central China:

784 an in situ SIMS study with implications for the source of sulfur.

785 Mineralium Deposita, 50, 643–656.

786 Chen, Y.J., Pirajno, F., and Qi, J.P. (2005) Origin of gold metallogeny and

787 sources of ore-forming fluids, Jiaodong Province, Eastern China.

788 International Geology Review, 47, 530–549.

789 Dare, S.A.S., Barnes, S.J., Beaudoin, G., Méric, J., Boutroy, E., and Potvin-

790 Doucet, C. (2014) Trace elements in magnetite as petrogenetic indicators.

791 Mineralium Deposita, 49, 785–796.

792 Deditius, A.P., Reich, M., Kesler, S.E., Utsunomiya, S., Chryssoulis, S.L.,

793 Walshe, J., and Ewing, R.C. (2014) The coupled geochemistry of Au and

794 As in pyrite from hydrothermal ore deposits. Geochimica et

795 Cosmochimica Acta, 140, 644–670.

796 Deng, J., Wang, C., Bagas, L., Carranza, E.J.M., and Lu, Y. (2015) Cretaceous–

797 Cenozoic tectonic history of the Jiaojia Fault and gold mineralization in

798 the Jiaodong Peninsula, China: constraints from zircon U-Pb, illite K-Ar,

799 and apatite fission track thermochronometry. Mineralium Deposita, 50,

800 987–1006.

801 Deng, J., Qiu, K.F., Wang, Q.F., Goldfarb, R.J., Yang, L.Q., Zi, J.W., Geng, J.Z.,

802 and Ma, Y. (2020a) In-situ dating of hydrothermal monazite and

803 implications on the geodynamic controls of ore formation in the Jiaodong

804 gold province, eastern China. Economic Geology, 115, 671–685.

805 Deng, J., Yang, L.Q., Groves, D.I., Zhang, L., Qiu, K.F., and Wang, Q.F. (2020b)

806 An integrated mineral system model for the gold deposits of the giant

807 Jiaodong province, eastern China. Earth-Science Reviews, 208, 103274.

808 Dora, M.L., Upadhyay, D., Randive, K.R., Shareef, M., Baswani, S.R., and

809 Ranjan, S. (2020) Trace element geochemistry of magnetite and pyrite and

810 sulfur isotope geochemistry of pyrite and barite from the Thanewasna Cu-

811 (Au) deposit, western Bastar Craton, central India: Implication for ore

812 genesis. Ore Geology Reviews, 117, 103262.

813 Fan, H.R., Zhai, M.G., Xie, Y.H., and Yang, J.H. (2003) Ore-forming fluids

814 associated with granite-hosted gold mineralization at the Sanshandao

815 deposit, Jiaodong gold province, China. Mineralium Deposita, 38, 739–

816 750.

817 Fan, H.R., Hu, F.F., Yang, J.H., and Zhai, M.G. (2007) Fluid evolution and

818 large-scale gold metallogeny during Mesozoic tectonic transition in the

819 Jiaodong Peninsula, eastern China. Geological Society London Special

820 Publications, 280, 303–316.

821 Feng, K., Fan, H.R., Hu, F.F., Yang, K.F., Liu, X., Shangguan, Y.N., and Jiang,

822 P. (2018) Involvement of anomalously As-Au-rich fluids in the

823 mineralization of the Heilan’gou gold deposit, Jiaodong, China: evidence

824 from trace element mapping and in-situ sulfur isotope composition.

825 Journal of Asian Earth Sciences, 160, 304–321.

826 Feng, K., Fan, H.R., Groves, D.I., Yang, K.F., Hu, F.F., Liu, X., and Cai, Y.C.

827 (2020) Geochronological and sulfur isotopic evidence for the genesis of

828 the post-magmatic, deeply sourced, and anomalously gold-rich Daliuhang

829 orogenic deposit, Jiaodong, China. Mineralium Deposita, 55, 293–308.

830 Fiege, A., Holtz, F., Shimizu, N., Mandeville, C.W., Behrens, H., and Knipping,

831 J.L. (2014) Sulfur isotope fractionation between fluid and andesitic melt:

832 An experimental study. Geochimica et Cosmochimica Acta, 142, 501–521.

833 Frost, B.R., and Lindsley, D.H. (1992) Equilibria among Fe-Ti oxides,

834 pyroxenes, olivine, and quartz: part II. Application. American Mineralogist,

835 77, 1004–1020.

836 Frost, B.R., Chamberlain, K.R., and Schumacher, J.C. (2001) Sphene (titanite):

837 phase relations and role as a geochronometer. Chemical Geology, 172,

838 131–148.

839 Fu, J.L., Hu, Z.C., Zhang, W., Yang, L., Liu, Y.S., Li, M., Zong, K.Q., Gao, S.,

840 and Hu, S.H. (2016) In situ, sulfur isotopes (δ34S and δ33S) analyses in

841 sulfides and elemental sulfur using high sensitivity cones combined with

842 the addition of nitrogen by Laser Ablation MC-ICP-MS. Analytica

843 Chimica Acta, 911, 14–26.

844 Gilbert, S.E., Danyushevsky, L.V., Rodemann, T., Shimizu, N., Gurenko, A.,

845 Meffre, S., Thomas, H., Large, R.R., and Death, D. (2014) Optimisation of

846 laser parameters for the analysis of sulphur isotopes in sulphide minerals

847 by laser ablation ICP-MS. Journal of Analytical Atomic Spectrometry, 29,

848 1042–1051.

849 Goldfarb, R.J., and Santosh, M. (2014) The dilemma of the Jiaodong gold

850 deposits: Are they unique? Geoscience Frontiers, 5, 139–153.

851 Goss, S.C., Wilde, S.A., Wu, F.Y., and Yang, J.H. (2010) The age, isotopic

852 signature and significance of the youngest Mesozoic granitoids in the

853 Jiaodong Terrain, Shandong Province, North China Craton. Lithos, 120,

854 309–326.

855 Groves, D.I., Goldfarb, R.J., Gebre-Mariam, M., Hagemann, S.G., and Robert,

856 F. (1998) Orogenic gold deposits: a proposed classification in the context

857 of their crustal distribution and relationship to other gold deposit types.

858 Ore Geology Reviews, 13, 7–27.

859 Hamilton, P.J., O’Nions, R.K., Bridgwater, D., and Nutman, A. (1983) Sm-Nd

860 studies of Archaean metasediments and metavolcanics from West

861 Greenland and their implications for the Earth’s early history. Earth and

862 Planetary Science Letters, 62, 263–272.

863 Hayden, L.A., Watson, E.B., Wark, D.A. (2008) A thermobarometer for sphene

864 (titanite). Contributions to Mineralogy and Petrology, 155, 529–540.

865 Hu, Z.C., Zhang, W., Liu, Y.S., Gao, S., Li, M., Zong, K.Q., Chen, H.H., and

866 Hu, S.H. (2015) “Wave” signal-smoothing and mercury-removing device

867 for laser ablation quadrupole and multiple collector ICPMS analysis:

868 application to lead isotope analysis. Analytical Chemistry, 87, 1152–1157.

869 Jacobsen, S.B., and Wasserburg, G.J. (1980) Sm-Nd isotopic evolution of

870 chondrites. Earth and Planetary Science Letters, 50, 139–155.

871 Li, J., Xu, L.L., Bi, X.W., Tang, Y.Y., Sheng, X.Y., Yu, H.J., Liu, G., and Ma, R.

872 (2020) New titanite U–Pb and molybdenite Re–Os ages for a hydrothermal

873 vein-type Cu deposit in the Lanping Basin, Yunnan, SW China: constraints

874 on regional metallogeny and implications for exploration. Mineralium

875 Deposita. https://doi.org/10.1007/s00126-020-00973-x.

876 Li, J.W., Deng, X.D., Zhou, M.F., Liu, Y.S., Zhao, X.F., and Guo, J.L. (2010)

877 Laser ablation ICP–MS titanite U–Th–Pb dating of hydrothermal ore

878 deposits: a case study of the Tonglushan Cu–Fe–Au skarn deposit, SE

879 Hubei Province, China. Chemical Geology, 270, 56–67.

880 Li, J.W., Bi, S.J., Selby, D., Chen, L., Vasconcelos, P., Thiede, D., Zhou, M.F.,

881 Zhao, X.F., Li, Z.K., and Qiu, H.N. (2012a) Giant Mesozoic gold

882 provinces related to the destruction of the North China craton. Earth and

883 Planetary Science Letters, 349, 26–37.

884 Li, L., Santosh, M., and Li, S.R. (2015) The ‘Jiaodong type’ gold deposits:

885 characteristics, origin and prospecting. Ore Geology Reviews, 65, 589–

886 611.

887 Li, R.C., Chen, H.Y., Xia, X.P., Yang, Q., Li, L., Xu, J., Huang, C., and

888 Danyushevsky, L.V. (2017) Ore fluid evolution in the giant Marcona Fe-

889 (Cu) deposit, Perú: Evidence from in-situ sulfur isotope and trace element

890 geochemistry of sulfides. Ore Geology Reviews, 86, 624–638.

891 Li, X.C., Fan, H.R., Santosh, M., Hu, F.F., Yang, K.F., Lan, T.G., Liu, Y., and

892 Yang, Y.H. (2012b) An evolving magma chamber within extending

893 lithosphere: an integrated geochemical, isotopic and zircon U–Pb

894 geochronological study of the Gushan granite, eastern North China Craton.

895 Journal of Asian Earth Sciences, 50, 27–43.

896 Li, X.H., Fan, H.R., Yang, K.F., Hollings, P., Liu, X., Hu, F.F., and Cai, Y.C.

897 (2018) Pyrite textures and compositions from the Zhuangzi Au deposit,

898 southeastern North China craton: Implication for ore-forming processes.

899 Contributions to Mineralogy and Petrology, 73, 73.

900 Li, X.H., Fan, H.R., Hu, F.F., Hollings, P., Yang, K.F., and Liu, X. (2019a)

901 Linking lithospheric thinning and magmatic evolution of late Jurassic to

902 early cretaceous granitoids in the Jiaobei Terrane, southeastern North

903 China Craton. Lithos, 324, 280–296.

904 Li, X.H., Fan, H.R., Santosh, M., Yang, K.F., Peng, H.W., Hollings, P., and Hu,

905 F.F. (2019b) Mesozoic felsic dikes in the Jiaobei Terrane, southeastern

906 North China Craton: Constraints from zircon geochronology and

907 geochemistry, and implications for gold metallogeny. Journal of

908 Geochemical Exploration, 201, 40–55.

909 Liang, Y.Y., Deng, J., Liu, X., Wang, Q., Ma, Y., Gao, T., Zhao, L. (2019) Water

910 contents of Early Cretaceous mafic dikes in the Jiaodong Peninsula,

911 eastern North China Craton: insights into an enriched lithospheric mantle

912 source metasomatized by paleo–Pacific Plate subduction–related fluids.

913 The Journal of Geology, 127, 343–362.

914 Liu, S., Hu, R.Z., Gao, S., Feng, C.X., Yu, B.B., Qi, Y.Q., Wang, T., Feng, G.Y.,

915 and Coulson, I.M. (2009) Zircon U-Pb age, geochemistry and Sr–Nd–Pb

916 isotopic compositions of adakitic volcanic rocks from Jiaodong, Shandong

917 Province, Eastern China: constraints on petrogenesis and implications.

918 Journal of Asian Earth Sciences, 35, 445–458.

919 Liu, Y.S., Hu, Z.C., Gao, S., Günther, D., Xu, J., Gao, C.G., and Chen, H.H.

920 (2008) In situ analysis of major and trace elements of anhydrous minerals

921 by LA-ICP-MS without applying an internal standard. Chemical Geology,

922 257, 34–43.

923 Ludwig, K.R. (2003) ISOPLOT 3.00: A Geochronological Toolkit for Microsoft

924 Excel. Berkeley Geochronology Center, California, Berkeley, pp 39.

925 Ma, W.D., Fan, H.R., Liu, X., Pirajno, F., Hu, F.F., Yang, K.F., Yang, Y.H., Xu,

926 W.G., and Jiang, P. (2017) Geochronological framework of the Xiadian

927 gold deposit in the Jiaodong province, China: implications for the timing

928 of gold mineralization. Ore Geology Reviews, 86, 196–211.

929 Mao, J.W., Wang, Y.T., Li, H.M., Pirajno, F., Zhang, C.Q., and Wang, R.T. (2008)

930 The relationship of mantle-derived fluids to gold metallogenesis in the

931 Jiaodong Peninsula: evidence from D-O-C-S isotope systematics. Ore

932 Geology Reviews, 33, 361–381.

933 Miao, L.C., Luo, Z.K., Huang, J.Z., Guan, K., Wang, L.G., McNaughton, N.J.,

934 and Groves, D.I. (1997) Zircon Sensitive High Resolution Ion Microprobe

935 (SHRIMP) study of granitoid intrusions in Zhaoye Gold Belt of Shandong

936 Province and its implication. Science in China (Series D), 40, 361–369.

937 Mills, S.E., Tomkins, A.G., Weinberg, R.F., and Fan, H.R. (2015) Implications

938 of pyrite geochemistry for gold mineralisation and remobilisation in the

939 Jiaodong gold district, northeast China. Ore Geology Reviews, 71, 150–

940 168.

941 Morad, S., El-Ghali, M.A.K., Caja, M.A., Al-Ramadan, K., and Mansurbeg, H.

942 (2009) Hydrothermal alteration of magmatic titanite: evidence from

943 Proterozoic granitic rocks, Southeastern Sweden. Canadian Mineralogist,

944 47, 801–811.

945 Nie, F.J., Jiang, S.H., and Liu, Y. (2004) Intrusion-related gold deposits of North

946 China Craton, People's Republic of China. Resource Geology, 54, 299–

947 324.

948 Ohmoto, H. (1972) Systematics of sulfur and carbon isotopes in hydrothermal

949 ore deposits. Economic Geology, 67, 551–578.

950 Olin, P.H., and Wolff, J.A. (2012) Partitioning of rare earth and high field

951 strength elements between titanite and phonolitic liquid. Lithos, 128–131,

952 46–54.

953 Peterson, E., and Mavrogenes, J. (2014) Linking high-grade gold mineralization

954 to earthquake-induced fault-valve processes in the Porgera gold deposit,

955 Papua New Guinea. Geology, 42, 383–386.

956 Piccoli, P., Candela, P., and Rivers, M. (2000) Interpreting magmatic processes

957 from accessory phases: titanite–a small-scale recorder of large-scale

958 processes. Transactions of the Royal Society of Edinburgh Earth Sciences,

959 91, 257–267.

960 Pokrovski, G.S., Kara, S., and Roux, J. (2002) Stability and solubility of

961 arsenopyrite, FeAsS, in crustal fluids. Geochimica et Cosmochimica Acta,

962 66, 2361–2378.

963 Pokrovski, G.S., Borisova, A.Y., and Bychkov, A.Y. (2013) Speciation and

964 transport of metals and metalloids in geological vapors: Reviews in

965 Mineralogy and Geochemistry, 76, 165–218.

966 Qiu, Y.M., Groves, D.I., McNaughton, R.J., and Phillips, G.N. (2002) Nature,

967 age, and tectonic setting of granitoid-hosted, orogenic gold deposits of the

968 Jiaodong Peninsula, eastern North China Craton, China. Mineraliun

969 Deposita, 37, 283–305.

970 Shen, Y.K., Guo, T., Yang, Y.Q., Chen, Z.L., Wei, C.S., and Sun, H.S. (2016)

971 Discovery of biotite monzonite and Ar-Ar thermochronology significance

972 in Linglong gold field. Journal of Geomechanics, 22, 778–793 (in Chinese

973 with English abstract).

974 Sillitoe, R.H., and Thompson, J.F. (1998) Intrusion–related vein gold deposits:

975 types, tectono-magmatic settings and difficulties of distinction from

976 orogenic gold deposits. Resource Geology, 48, 237–250.

977 Spandler, C., Hammerli, J., Sha, P., Hilbert-Wolf, H., Hu, Y., Roberts, E., and

978 Schmitz, M. (2016) MKED1: A new titanite standard for in situ analysis

979 of Sm–Nd isotopes and U–Pb geochronology. Chemical Geology, 425,

980 110–126.

981 Sun, W.D., Ding, X., Hu, Y.H., and Li, X.H. (2007) The golden transformation

982 of the Cretaceous plate subduction in the west Pacific. Earth and Planetary

983 Science Letters, 262, 533–542.

984 Tanner, D., Henley, R.W., Mavrogenes, J.A., and Holden, P. (2016) Sulfur

985 isotope and trace element systematics of zoned pyrite crystals from the El

986 Indio Au–Cu–Ag deposit, Chile. Contributions to Mineralogy and

987 Petrology, 171, 33.

988 Tiepolo, M., Oberti, R., and Vannucci, R. (2002) Trace-element incorporation

989 in titanite: constraints from experimentally determined solid/liquid

990 partition coefficients. Chemical Geology, 191, 105–119.

991 Wang, Z.C., Cheng, H., Zong, K.Q., Geng, X.L., Liu, Y.S., Yang, J.H., Wu, F.Y.,

992 Becker, H., Foley, S., and Wang, C.Y. (2020) Metasomatized lithospheric

993 mantle for Mesozoic giant gold deposits in the North China craton.

994 Geology, 48, 169–173.

995 Wen, B.J., Fan, H.R., Santosh, M., Hu, F.F., Pirajno, F., Yang, K.F. (2015)

996 Genesis of two different types of gold mineralization in the Linglong gold

997 field, China: constrains from geology, fluid inclusions and stable isotope.

998 Ore Geology Reviews, 65, 643–658.

999 Wiedenbeck, M., Alle, P., Corfu, F., Griffin, W.L., Meier, M., Oberli, F., Quadt,

1000 A.V., Roddick, J.C., and Spiegel, W. (1995) Three natural zircon standards

1001 for U–Th–Pb, Lu–Hf, trace element and REE analyses. Geostandards and

1002 Geoanalytical Research, 19, 1–23.

1003 Wu, F.Y., Yang, J.H., Xu, Y.G., Wilde, S.A., and Walker, R.J. (2019) Destruction

1004 of the North China craton in the Mesozoic. Annual Review of Earth and

1005 Planetary Sciences, 47, 173–195.

1006 Xiao, X., Zhou, T.F., White, N.C., Zhang, L.J., Fan, Y., and Chen, X.F. (2021)

1007 Multiple generations of titanites and their geochemical characteristics

1008 record the magmatic-hydrothermal processes and timing of the

1009 Dongguashan porphyry-skarn Cu-Au system, Tongling district, Eastern

1010 China. Mineralium Deposita, 56, 363–380.

1011 Xie, L., Wang, R.C., Chen, J., and Zhu, J.C. (2010) Mineralogical evidence for

1012 magmatic and hydrothermal processes in the Qitianling oxidized tin

1013 bearing granite (Hunan, South China): EMP and (MC)-LA-ICPMS

1014 investigations of three types of titanite. Chemical Geology, 276, 53–68.

1015 Xie, S.W., Wu, Y.B., Zhang, Z.M., Qin, Y.C., Liu, X.C., Wang, H., Qin, Z.W.,

1016 Liu, Q., and Yang, S.H. (2012) U-Pb ages and trace elements of detrital

1017 zircons from Early Cretaceous sedimentary rocks in the Jiaolai Basin,

1018 north margin of the Sulu UHP terrain: provenances and tectonic

1019 implications. Lithos, 154, 346–360.

1020 Xu, H.J., Zhang, J.F., Wang, Y.F., and Liu, W.L. (2016) Late triassic alkaline

1021 complex in the Sulu UHP terrane: implications for post-collisional

1022 magmatism and subsequent fractional crystallization. Gondwana Research,

1023 35, 390–410.

1024 Xu, L., Hu, Z., Zhang, W., Yang, L., Liu, Y., Gao, S., Luo, T., and Hu, S. (2015)

1025 In situ Nd isotope analyses in geological materials with signal

1026 enhancement and non-linear mass dependent fractionation reduction using

1027 laser ablation MC-ICP-MS. Journal of Analytical Atomic Spectrometry, 30,

1028 232–244.

1029 Yang, J.H., and Zhou, X.H. (2001) Rb-Sr, Sm-Nd, and Pb isotope systematics

1030 of pyrite: Implications for the age and genesis of lode gold deposits.

1031 Geology, 29, 711–714.

1032 Yang, K.F., Fan, H.R., Santosh, M., Hu, F.F., Wilde, S.A., Lan, T.G., Lu, L.N.,

1033 and Liu, Y.S. (2012) Reactivation of the Archean lower crust: implications

1034 for zircon geochronology, elemental and Sr-Nd-Hf isotopic geochemistry

1035 of late Mesozoic granitoids from northwestern Jiaodong Terrain, the North

1036 China Craton. Lithos, 146, 112–127.

1037 Yang, K.F., Jiang, P., Fan, H.R., Zuo, Y.B., and Yang, Y.H. (2018) Tectonic

1038 transition from a compressional to extensional metallogenic environment

1039 at ∼120 Ma revealed in the Hushan gold deposit, Jiaodong, North China

1040 Craton. Journal of Asian Earth Sciences, 160, 408–425.

1041 Zhai, M.G., Fan, H.R., Yang, J.H., and Miao, L.C. (2004) Large-scale cluster of

1042 gold deposits in east Shandong: anorogenic metallogenesis. Earth Science

1043 Frontiers, 11, 85–98 (in Chinese with English abstract).

1044 Zhang, L., Weinberg, R.F., Yang, L.Q., Groves, D.I., Sai, S.X., Matchan, E.,

1045 Phillips, D., Kohn, B.P., Miggins, D.P., Liu, Y., and Deng, J. (2020)

1046 Mesozoic orogenic gold mineralization in the Jiaodong Peninsula, China:

1047 a focused event at 120 ± 2 Ma during cooling of pre-gold granite intrusions.

1048 Economic Geology, 115, 415–441.

1049 Zheng, Y.F. (2008) A perspective view on ultrahigh-pressure metamor- phism

1050 and continental collision in the Dabie–Sulu orogenic belt. Chinese Science

1051 Bulletin, 53, 3081–3104.

1052 Zhi, Z.Y., Li, L., Li, S.R., Santosh, M., Yuan, M.W., and Alam, M. (2019)

1053 Magnetite as an indicator of granite fertility and gold mineralization: A

1054 case study from the Xiaoqinling gold province, North China Craton. Ore

1055 Geology Reviews, 115, 103159.

1056 Zhou, J.B., Wilde, S.A., Zhao, G.C., Zhang, X.Z., Zheng, C.Q., Jin, W., and

1057 Cheng, H. (2008) SHRIMP U–Pb zircon dating of the Wulian complex:

1058 defining the boundary between the North and South China Cratons in the

1059 Sulu Orogenic Belt, China. Precambrian Research, 162, 559–576.

1060 Zhou, T.H., Goldfarb, R.J., and Phillips, G.N. (2002) Tectonics and distribution

1061 of gold deposits in China: an overview. Mineralium Deposita, 37, 249–282.

1062 Zhu, R.X., Fan, H.R., Li, J.W., Meng, Q.R., Li, S.R., and Zeng, Q.D. (2015)

1063 Decratonic gold deposits. Science China Earth Sciences, 58, 1523–1537.

1064

1065 Figure captions

1066 Fig. 1 Geological map showing the distribution of basement rocks, UHP

1067 metamorphic rocks, Mesozoic igneous rocks and Au deposits in the three

1068 Au belts of the Jiaodong Au province. Modified after Yang et al. (2012).

1069 Fig. 2 (a) Geologic map of the Linglong goldfield, showing the distribution of

1070 the orebodies, modified after Wen et al. (2015). (b) Cross-section of the

1071 Line 72 in the Linglong goldfield and sample location.

1072 Fig. 3 Photographs of hand specimen and thin section cross-polarized

1073 illumination of biotite monzonite, containing disseminated pyrite. (a, c)

1074 Fresh-weakly altered biotite monzonite. (b, d) Silicified-sericitized biotite

1075 monzonite. (e) Quartz-pyrite veins in Luanjiahe granite. (f) Mafic enclave

1076 in Gushan granite. (g, h) Photomicrographs of biotite monzonite. Qtz

1077 quartz, Kfs k-feldspar, Pl plagioclase, Bt biotite, Cal calcite, Py pyrite.

1078 Fig. 4 BSE images of three types of titanite and monazite. (a, b) Ttn1 with

1079 euhedral-subhedral shape and oscillatory zones. (c-e) Ttn2 with euhedral-

1080 subhedral shape and core-rim texture, coexisting with pyrite and pyrrhotite.

1081 (f-h) Dark-grey BSE Ttn3 penetrating and/or overgrowing Ttn2 and

1082 coexisting with calcite. (i) Hydrothermal monazite for U-Pb dating, from

1083 quartz-sulfide veins. Red circles show the location of points for LA-

1084 ICPMS U-Pb dating, together with 207Pb-corrected 206Pb/238U ages. Ttn

1085 titanite, Py pyrite, Po pyrrhotite, Mt magnetite, Cal calcite, Bt biotite, Qtz

1086 quartz, Mnz monazite, Rt rutile.

1087 Fig. 5 Selected binary plots showing the geochemical characteristics of three

1088 types of titanite. Plots of (a) TiO2 versus FeO, (b) TiO2 versus MnO, (c)

1089 CaO versus REE, (d) CaO versus LREE/HREE, (e) EuN/EuN* versus

1090 Lu/Hf, (f) Th versus Pb.

1091 Fig. 6 Tera-Wasserburg diagrams and corresponding weighted mean plots of

1092 age data for (a) magmatic titanite (Ttn1), (b) hydrothermal titanite (Ttn2,

1093 Ttn3) from the biotite monzonite, and (c) hydrothermal monazite from the

1094 quartz-pyrite veins.

1095 Fig. 7 εNd(t) values of titanite from the biotite monzonite, and comparison with

1096 whole-rock εNd(t) values of Mesozoic magmatic rocks in the Jiaodong

1097 Peninsula. Data are from Li et al. (2012); Yang et al. (2012); Cai et al.

1098 (2013); Li et al. (2019); and this study.

1099 Fig. 8 Reflected light photomicrographs and BSE images of four types of pyrite.

1100 (a, e) Py1 with homogeneous texture under BSE is from the fresh to

1101 weakly altered biotite monzonite, coexisting with magnetite, enclosing

1102 magmatic titanite and plagioclase. (b, f) Py2 occurs with pyrrhotite and

1103 shows heterogeneous texture under BSE, from silicified-sericitized biotite

1104 monzonite. (c, g) Py3 is from the quartz-pyrite veins, showing systematic

1105 zones. (d, h) Py4 is from the mafic enclave in the Gushan granite,

1106 occurring with magnetite, chalcopyrite and allanite. Py pyrite, Po

1107 pyrrhotite, Mt magnetite, Ccp chalcopyrite, Ttn titanite, Pl plagioclase,

1108 Mnz monazite, Aln allanite.

1109 Fig. 9 Trace elements of four types of pyrite. Note that some data with a value

1110 of 0 are not plotted on the diagram, and the contents of Sb and Au for Py4

1111 are below the detection limit.

1112 Fig. 10 In situ sulfur isotopes of four types of pyrite and pyrrhotite. The d34S

1113 peak of pyrite from the Jiaodong Au province is from Mao et al. (2008),

1114 Feng et al. (2018), Li et al. (2018), and Deng et al. (2020b).

1115 Fig. 11 Schematic diagram showing (a) the formation of the biotite monzonite,

1116 and the contribution of magmatic-hydrothermal fluids to Au deposits in the

1117 Jiaobei Terrane; (b) titanite and pyrite fingerprint of hydrothermal fluids

1118 exsolved from the monzonitic magma. The monzonite formed by partial

1119 melting of the lower crust of the NCC with involvement of mantle

1120 materials at ca. 120 Ma, probably leading to preconcentration of Au,

1121 releasing auriferous magmatic-hydrothermal fluids at late stage.

1122

1123 Table Captions

1124 Table 1 Results of in situ LA-ICPMS U-Pb dating on titanite and monazite

1125

1126 Electronic Supplementary materials

1127 Supplementary Table 1 In situ major, trace elements and Nd isotopes of

1128 magmatic and hydrothermal titanite

1129 Supplementary Table 2 In situ trace elements and/or sulfur isotopes of pyrite,

1130 pyrrhotite and magnetite

1131 Supplementary Figures Including Fig. A1 Chondrite-normalized REE

1132 distribution patterns for three types of titanite from the biotite monzonite;

1133 Fig. A2 Discrimination of magnetite from the monzonite based on Ti vs.

1134 Ni/Cr, after Dare et al. (2014)

Always consult and cite the final, published document. See http:/www.minsocam.org or GeoscienceWorld Table 1 Results of in-situ LA-ICPMS U-Pb dating on titanite and monazite

207 Spot Isotopic ratios Pb-corr. Age/Ma No. 207Pb/206Pb 1σ 207Pb/235U 1σ 206Pb/238U 1σ rho 206Pb/238U 1σ Magmatic Ttn1 1 0.1598 0.0208 0.3634 0.0221 0.0211 0.0008 0.61 116.1 4.9

2 0.2606 0.0235 0.6647 0.0403 0.0234 0.0009 0.65 110.1 5.8

3 0.1633 0.0274 0.6964 0.0738 0.0234 0.0019 0.75 128.3 11.8

4 0.0729 0.0044 0.1717 0.0086 0.0182 0.0005 0.50 112.7 2.9

5 0.2633 0.0319 0.6636 0.0451 0.0243 0.0014 0.86 113.7 8.9

6 0.0766 0.0048 0.1939 0.0102 0.0192 0.0004 0.42 118.5 2.7

7 0.1197 0.0086 0.3036 0.0164 0.0208 0.0006 0.54 120.9 3.9

8 0.0627 0.0024 0.1663 0.0058 0.0195 0.0003 0.44 122.4 1.9

9 0.0921 0.0056 0.2367 0.0107 0.0197 0.0004 0.47 119.1 2.7

10 0.2415 0.0223 0.7310 0.0459 0.0243 0.0012 0.77 117.7 7.4

11 0.1194 0.0243 1.1518 0.1443 0.0287 0.0027 0.76 166.9 17.1

12 0.0866 0.0265 7.4754 1.3708 0.0813 0.0131 0.88 485.7 78.0

13 0.4009 0.0605 33.8111 3.6589 0.2907 0.0282 0.89 1002.1 140.6

14 0.2696 0.0315 6.4758 0.4805 0.0707 0.0051 0.98 323.7 30.9

15 0.2094 0.0407 4.5784 0.6724 0.0599 0.0062 0.70 303.0 37.5

16 0.2757 0.0431 16.8791 1.9652 0.1484 0.0133 0.77 662.4 74.4

17 0.1520 0.0297 2.6638 0.2484 0.0382 0.0033 0.92 211.3 20.4

18 0.3438 0.0383 4.1160 0.2788 0.0578 0.0037 0.95 232.0 22.6

Hydrothermal Ttn2 and Ttn3 1 0.2123 0.0221 0.4920 0.0286 0.0224 0.0010 0.79 114.0 6.5

2 0.2795 0.0168 0.8772 0.0412 0.0249 0.0008 0.68 113.5 5.0

3 0.1340 0.0066 0.3677 0.0143 0.0208 0.0005 0.58 118.8 2.9

4 0.1493 0.0062 0.4139 0.0138 0.0208 0.0004 0.60 116.1 2.6

5 0.3672 0.0403 1.6644 0.1172 0.0315 0.0019 0.86 121.6 12.0

6 0.3558 0.0225 1.1918 0.0577 0.0279 0.0012 0.86 110.0 7.3

7 0.4104 0.0244 1.7262 0.0935 0.0328 0.0011 0.60 115.3 6.7

8 0.1584 0.0059 0.4712 0.0157 0.0220 0.0005 0.66 121.2 3.0

9 0.2281 0.0108 0.7052 0.0224 0.0243 0.0006 0.80 120.7 3.9

10 0.2814 0.0174 0.9036 0.0474 0.0259 0.0009 0.63 117.3 5.4

11 0.4367 0.0301 1.9547 0.1623 0.0330 0.0018 0.66 109.2 11.3

12 0.4134 0.0277 1.5193 0.0632 0.0318 0.0012 0.91 111.0 7.5

13 0.1636 0.0086 0.4884 0.0188 0.0229 0.0005 0.61 125.1 3.4

14 0.2364 0.0104 0.7566 0.0281 0.0244 0.0007 0.72 119.6 4.1

15 0.2455 0.0192 0.6450 0.0320 0.0222 0.0008 0.69 107.2 4.8

16 0.4707 0.0257 1.9566 0.0717 0.0333 0.0010 0.86 101.1 6.5

17 0.4220 0.0347 1.2847 0.0570 0.0290 0.0012 0.89 99.4 7.2

18 0.4063 0.0502 3.0899 0.2263 0.0416 0.0026 0.86 147.4 16.2

19 0.6842 0.0661 3.9290 0.1787 0.0524 0.0024 0.99 70.4 14.5

Hydrothermal monazite 1 0.1665 0.0250 0.4325 0.0443 0.0216 0.0012 0.55 117.3 7.6

2 0.1043 0.0149 0.2680 0.0294 0.0203 0.0007 0.30 120.6 4.2

3 0.5016 0.0710 2.0192 0.2391 0.0380 0.0033 0.73 102.8 20.4

4 0.4416 0.0742 1.9995 0.2439 0.0360 0.0026 0.60 114.9 16.4

5 0.6665 0.0535 6.8691 0.4026 0.0806 0.0038 0.81 110.1 22.8

6 0.5684 0.0582 3.6402 0.3118 0.0542 0.0030 0.65 117.1 18.4

7 0.5088 0.0335 3.0583 0.1813 0.0454 0.0019 0.72 120.0 12.0

8 0.2305 0.0216 0.7186 0.0597 0.0241 0.0011 0.56 118.1 7.0

9 0.6125 0.0599 3.9060 0.2581 0.0598 0.0029 0.73 107.9 17.5

10 0.2241 0.0256 0.8614 0.1738 0.0234 0.0020 0.42 116.0 12.6

11 0.4499 0.0262 2.3441 0.1800 0.0387 0.0018 0.62 120.9 11.4

12 0.4165 0.0691 1.7802 0.4106 0.0365 0.0039 0.47 124.1 24.4

13 0.2326 0.0330 0.8270 0.1462 0.0261 0.0016 0.35 127.4 10.0

14 0.2546 0.0323 0.8865 0.1923 0.0253 0.0020 0.36 119.0 12.5

15 0.2728 0.0438 0.7743 0.0685 0.0248 0.0011 0.50 113.4 6.9

16 0.3956 0.0681 1.5283 0.1612 0.0322 0.0021 0.62 114.9 13.1

17 0.2882 0.0314 1.1070 0.1890 0.0273 0.0021 0.45 121.1 13.0

18 0.5109 0.0831 2.0165 0.2047 0.0396 0.0032 0.79 104.3 19.8

19 0.6057 0.1017 3.6096 0.3799 0.0574 0.0056 0.92 106.8 33.9